Abstract
Purpose
Sleep of sufficient duration and quality is crucial for physical and mental health and performance. In modern urban society, work-related factors such as perceived workload and limited sleep duration on workdays appear to have a significant impact on sleep and well-being. Our current study examines both subjective and objective measures of sleep on workdays and work-free, as well as their associations.
Methods
The study combines questionnaires and longitudinal wearable (Fitbit Inspire 2) sleep data from young, healthy adults (aged 18–48 years) in a German metropolitan region (n = 67). Associations were investigated by Spearman correlation analyses with a 95% confidence interval.
Results
Perceived workload was associated with symptoms of depression/anxiety, daytime dysfunction, as well as subjective sleep quality, in particular on workdays. Sleep and sleep stages were objectively longer and later on work-free days than on work-days. Likewise, objective sleep efficiency and subjective sleep quality were higher on work-free days. Longer sleep duration, thus sleep compensation, on work-free days was associated with subjectively more restful and better sleep, and objectively with later phase and a higher proportion of REM sleep.
Conclusion
These data suggest that work-related sleep loss should be avoided whenever possible or at least compensated for on work-free days in order to achieve not only sufficient sleep duration but also good sleep quality, which is pivotal for mental health and performance.
Keywords: REM sleep, chronotype, social jet lag, sleep efficiency, PHQ-4, workload, PSQI
Introduction
Restful sleep is essential for regeneration and energy balance and plays an important role for mental- and physical health.1–8 Insufficient sleep is manifested by daytime fatigue and reduced performance and can lead to increased stress and symptoms of anxiety and depression.1–8 Sleep quantity and quality are interrelated components of restful sleep.9 Both contribute equally to physical and mental well-being6,10–12 and are negatively associated with workload.11 In particular, short sleep duration on workdays appears to affect the subjective perception of workload, which in turn is associated with symptoms of depression.13
The required sleep duration for adults is somewhere between 7–9 hours,14 but shows large individual differences. In addition, there are both extrinsic and intrinsic factors15 that influence bedtime and thus also sleep duration and quality. Intrinsic components include the homeostatic sleep drive as well as a circadian component.16,17 The individual phase of the sleep propensity rhythm is essentially controlled by an internal clock18 and can also be used to determine the chronotype.19–22 Particularly in modern urban society,23 extrinsic factors include artificial light in the evening, which can delay bedtime, and the early start of work, which can be associated with premature end to sleep. Sleep deficits due to getting up early on workdays often cannot be counterbalanced simply by going to bed earlier. Due to the intrinsic phase of the sleep propensity rhythm, an earlier bedtime may just lead to a longer sleep latency if the homeostatic sleep pressure is not high enough. However, work-free days can be used to compensate for the accumulated sleep dept. Sleep compensation is probably one of the reasons why sleep on work-free days is perceived as much better than on workdays.24–27
Measuring objective sleep quality still presents a major challenge. Because sleep continuity is a major hallmark of sleep quality, the time spent awake after sleep onset (WASO) and the ratio of total sleep time to time spent in bed, which is referred to as sleep efficiency, provide qualitative measures. In addition, sleep architecture is also of particular importance for sleep quality. Each sleep cycle consists of a sequence of distinct sleep stages characterized by fluctuations in muscle tone, brain activity patterns, heart rate, body temperature, and eye movements.28 Each sleep cycle starts with brief transitions between wakefulness and light sleep (stage 1), followed by light sleep (stage 2) and deep sleep (stage 3) and ends with rapid eye movement (REM) sleep. Each cycle lasts roughly 90–120 minutes and repeats 4–6 times per night. Over the course of the night, the length of the stages changes, with deep sleep decreasing and REM sleep increasing. Deep sleep and REM sleep each account for about a quarter of total sleep.28 REM sleep appears to be particularly important for sleep quality.27,29,30 Since REM sleep occurs predominantly in the second half of the night and the early morning,31,32 it seems to be particularly affected by being woken up by the alarm clock in the early morning.27,33 Sleep stages can be precisely determined using polysomnography, which typically requires a visit to a sleep laboratory. However, they can also be assessed by algorithms based on heart rate and movement patterns, using wrist-worn wearables. Analysis of sleep and sleep stages such as deep sleep, WASO and REM sleep by Fitbit wearables is less accurate than polysomnography,34–39 but can be effectively used in field studies.27,33 Fitbit Alta HR and Inspire HR, the predecessors of Fitbit Inspire 2 used in our study, shows comparatively good specificity and sensitivity for sleep and sleep stages.35,39 A meta-analysis shows that Fitbit wearables in particular exhibit a high degree of comparability with polysomnography in terms of key sleep parameters such as sleep duration, WASO, and sleep efficiency.40 Fitbit-derived sleep data have been used in epidemiological research, showing strong associations between sleep quality and health outcomes.41
Previous studies have examined differences in sleep quality between workdays and work-free days based on either standardized questionnaires24,25 or wearable devices,27 but not yet combined. Our hypothesis in this study was that there is an association between subjective sleep quality on working days and work-free days determined by questionnaire and the corresponding objectively measured longitudinal sleep data from a Fitbit wearable device. Our particular focus is on better understanding the differences of key sleep parameters such as sleep duration, WASO, sleep efficiency and REM sleep between workdays and work-free days as a basis for improving sleep and thus performance and well-being.
Materials and Methods
Procedure and Participants
Part of the data used for this study came from the same dataset as previously described.24 The participants were recruited in the Düsseldorf metropolitan region between 5 May 2023 and 18 September 2024 through personal contacts using a snowball system. Criteria for recruitment included age between 18 and 50, regular weekly working hours (including school/university), no shift work and no self-reported chronic illness (including sleep disorders).
All participants were informed about the study conditions and data protection regulations and gave their consent before study enrolment. Participants were equipped with Fitbit Inspire 2 multisensory (movement and heart rate) sleep trackers and asked to wear them as continuously as possible for three months, especially at night. The data collection period includes 64 days of the week, defined as working days, and 26 weekend days, defined as non-working days. After the data collection period, participants were asked to complete the online questionnaire where all questions were mandatory and could only be answered once. They were then automatically redirected to the Fitbit website, where they could authorize the transfer of activity and sleep data to our study server.
The study was conducted in accordance with the ethical requirements of the Declaration of Helsinki and approved by the Research Ethics Committee of the Faculty of Medicine of Heinrich Heine University (file number of approval: 2019–3786).
Questionnaire
The online questionnaire contained questions addressing general characteristics such as biological sex, age, body weight, height, the inclusion criteria and lifestyle factors.
The questions on work asked whether the participants had regular employment, whether they worked on weekends, whether they worked shifts, and how high they generally rate their workload on a four-point scale (low=0, moderate=1, high=2, very high=3).13,27
The questions about health included the 4-item patient health questionnaire for depression and anxiety (PHQ-4), according to Lowe et al (2010),42 each scored on a four-point scale (not at all=0, on single days=1, on more than half of the days=2, almost every day=3).
Self-reported sleep quality was assessed with a question on how tired participants felt when waking in the morning on a three-point scale (rested=0, tired=1, very tired=2)24 and with the Pittsburgh Sleep Quality Index (PSQI).30 These were asked separately for workdays and work-free days.24,25 The seven PSQI components, each assessed on a four-point scale24 and reflecting a different dimension of sleep, were summed to produce the global scores (range 0–21) and also considered separately. Lower scores indicate better sleep.
The consumption of alcohol and caffeinated drinks, as well as physical activity, were rated on a five-point scale (none = 0, 1–2 times per week = 1, several times per week = 2, once daily = 3, several times daily = 4), while drug use and consumption of nicotine (=smoking cigarettes/e-cigarettes) were rated on a two-point scale (no = 0, yes = 1).
Assessment of Activity, Sleep and Sleep Stages Based on Fitbit Wearable Data
Activity and sleep were assessed by Fitbit analyses over a period of three month. Proprietary Fitbit algorithms detect activity,43 sleep and sleep stages44 based on heartrate and activity patterns. A python-based custom software application was used to process Fitbit data. The variables were determined for the entire period (total) as well as separately for workdays and work-free days. The phases of total sleep and sleep stages were assessed based on their midpoint in local time (digital).
Data Analysis
Statistical analysis was performed using Prism Version 7.01 (GraphPad) and R studio Version 2023.12.0+369. Normal distribution of data was tested by D’Agostino and Pearson normality test. For descriptive statistics, the mean ± standard error of the mean was used for normally distributed data while the median with the interquartile range was used for non-normally distributed data.
Because not all variables were normally distributed, non-parametric tests were used consistently. Statistical significance was assumed at P<0.05. The Wilcoxon matched-pairs signed rank test was used to compare variables between workdays and work-free days. The strength and direction of relationships between categorial and continuous variables were analyzed by Spearman’s rank correlation analysis with a 95% confidence interval (CI). The strength and direction of relationships between a “yes/no” variable and continuous variables were analyzed by Point-biserial correlation with a 95% CI. The Holm-Bonferroni correction was used to reduce the probability of Type I errors in multiple correlations. Linear regression was used for graphical visualization of relationships, even if data were not normally distributed.
Results
Sample Size
Of the 75 subjects we recruited, we had to exclude 3 subjects because they met at least one exclusion criterion. 5 subjects were not included because they did not reach the cut-off of wearing the Fitbit tracker in 50% of the nights throughout the study period. This resulted in a total sample size of n=67. Based on this sample size and an estimated effect size of 0.4, a statistical power of 0.97 was achieved (G*Power).
General Characteristics, Depression/Anxiety and Workload Perception
The distribution of age, sex and BMI is summarized in Table 1. The median age was 25 (21–25) years, 39 were female and 28 male. The median BMI was 22.6 (21.2–25.6) kg/m2. The majority (61%) of participants were in the normal weight range (18.5–24.9 kg/m2) according to the WHO classifications.45
Table 1.
Sex, Age, and BMI (n=67)
| n | % of Total | |
|---|---|---|
| Sex | ||
| Female | 39 | 58 |
| Male | 28 | 42 |
| Age (years) | ||
| 18-27 | 42 | 63 |
| 28-37 | 16 | 24 |
| 38-47 | 9 | 13 |
| BMI (kg/m2) | ||
| <18.5 | 5 | 7 |
| 18.5–24.9 | 41 | 61 |
| 25.0–29.9 | 12 | 18 |
| 30.0–34.9 | 8 | 12 |
| 35.0–39.9 | 1 | 1 |
| ≥ 40.0 | 0 | 0 |
The distribution of the PHQ-4 score and the score for perceived workload is shown in Table 2. The majority of participants reported having a moderate (27%) to high (40%) workload. The majority (46%) of participants had a PHQ-4 score (depression/anxiety) in the normal range (0–2), followed by 39% in the mild range (3–5), according to Lowe et al (2010).42 Depression/anxiety and perception of workload were not significantly associated with general characteristics of the sample but with each other (Supplementary Table 1).
Table 2.
PHQ-4 (Depression/Anxiety) and Workload Perception (n=67)
| n | % of Total | |
|---|---|---|
| PHQ-4 score | ||
| 0-2 | 31 | 46 |
| 3-5 | 26 | 39 |
| 6-8 | 7 | 10 |
| 9-12 | 3 | 4 |
| Workload score | ||
| Low | 15 | 22 |
| Moderate | 18 | 27 |
| High | 27 | 40 |
| Very high | 7 | 10 |
Differences in Self-Reported and Wearable Measured Sleep Variables Between Workdays and Work-Free days
Table 3 summarizes the differences in self-reported sleep variables between workdays and work-free days. The global PSQI score was significantly higher on workdays than on work-free days, indicating poorer sleep on workdays. Consistently, the scores for tiredness on waking, C1 (subjective sleep quality), C2 (sleep latency), C3 (sleep duration), and C7 (daytime dysfunction) were significantly higher on workdays than on work-free days, indicating poorer sleep quality and quantity, and worse daytime performance on workdays. Tiredness on waking, subjective sleep quality and daytime dysfunction were associated with each other (Supplementary Table 2).
Table 3.
Differences in Self-Reported Sleep Variables Between Workdays and Work-Free days in the Sample (n=67)
| Median | IR | P value | |
|---|---|---|---|
| Tiredness on waking | 0.0005*** | ||
| Workdays | 2 | [1,2] | |
| Work-free days | 1 | [1,2] | |
| PSQI global | <0.0001**** | ||
| Workdays | 6 | [4,7] | |
| Work-free days | 4 | [1,2] | |
| PSQI components | |||
| C1 Subjective sleep quality | <0.0001**** | ||
| Workdays | 1 | [1,2] | |
| Work-free days | 1 | [0,1] | |
| C2 Sleep latency | 0.0035** | ||
| Workdays | 1 | [1,1] | |
| Work-free days | 1 | [0,1] | |
| C3 Sleep duration | <0.0001**** | ||
| Workdays | 1 | [1,2] | |
| Work-free days | 0 | [0,1] | |
| C4 Sleep efficiency | 0.16 | ||
| Workdays | 0 | [0,0] | |
| Work-free days | 0 | [0,0] | |
| C5 Sleep disturbances | 0.71 | ||
| Workdays | 1 | [1,1] | |
| Work-free days | 1 | [1,1] | |
| C6 Use of sleep medication | 0.08 | ||
| Workdays | 0 | [0,0] | |
| Work-free days | 0 | [0,0] | |
| C7 Daytime dysfunction | 0.004** | ||
| Workdays | 1 | [0,2] | |
| Work-free days | 1 | [0,1] |
Notes: Wilcoxon matched pairs signed rank test. **P<0.01, ***P<0.001, ****P<0.0001, higher values on work-free days.
Abbreviations: IR, interquartile range; PSQI, Pittsburgh Sleep Quality Index.
The differences in Fitbit-measured sleep variables between workdays and work-free days are shown in Table 4. The phases of total sleep, deep sleep REM sleep, and WASO were significantly later on work-free days than on workdays. In addition, the duration of total sleep, deep sleep REM sleep, and WASO were significantly longer on work-free days than on workdays. The proportion of REM sleep was significantly higher on work-free days than on workdays while the proportion of WASO was significantly lower on work-free days than on workdays. Sleep latency was lower and sleep efficiency was higher on work-free days. There was no significant difference in activity based on calculated calorie-consumption between workdays and work-free days (P=0.24, n=67).
Table 4.
Differences in Fitbit-Measured Sleep Variables Between Workdays and Work-Free days (n=67)
| Median | IR | P value | |
|---|---|---|---|
| Phase (digital local time) | |||
| Total sleep | 0.0001**** | ||
| Workdays | 4.2 | [3.1,4.9] | |
| Work-free days | 5.3 | [4.6,6.2] | |
| Deep sleep | 0.0001**** | ||
| Workdays | 2.6 | [1.6,3.9] | |
| Work-free days | 3.4 | [2.5,4.6] | |
| REM sleep | 0.0001**** | ||
| Workdays | 5.3 | [4.4,6] | |
| Work-free days | 6.2 | [5.2,7] | |
| WASO | 0.0001**** | ||
| Workdays | 4.3 | [3.1,5.4] | |
| Work-free days | 5.7 | [4.3,7] | |
| Duration (min) | |||
| Total sleep | 0.0001**** | ||
| Workdays | 446 | [412,487] | |
| Work-free days | 487 | [434,509] | |
| Deep sleep | 0.0001**** | ||
| Workdays | 70 | [63,80] | |
| Work-free days | 77 | [65,87] | |
| REM sleep | 0.0001**** | ||
| Workdays | 83 | [66,95] | |
| Work-free days | 87 | [74,148] | |
| WASO | 0.01* | ||
| Workdays | 56 | [51, 62] | |
| Work-free days | 58 | 52,66] | |
| Proportion (%) | |||
| Deep sleep | 0.48 | ||
| Workdays | 16 | [14,18] | |
| Work-free days | 16 | [14,18] | |
| REM sleep | 0.03* | ||
| Workdays | 18 | [16,20] | |
| Work-free days | 18 | [16,20] | |
| WASO | 0.008++ | ||
| Workdays | 12 | [12,14] | |
| Work-free days | 12 | [12,13] | |
| Sleep latency (min) | 0.0001++++ | ||
| Workdays | 6 | [4,7] | |
| Work-free days | 5 | [3, 6] | |
| Sleep efficiency (%) | 0.04* | ||
| Workdays | 88 | [86, 88] | |
| Work-free days | 89 | [87, 89] |
Notes: Wilcoxon matched pairs signed rank test. *P<0.05; ****P<0.0001, higher (later) values on work-free days. ++P<0.01; ++++P>0.0001, lower values on work-free days.
Abbreviations: IR, interquartile range; min, minutes; REM, rapid eye movement; WASO, wake after sleep onset.
Associations Among Subjective Sleep Quality, Daytime Dysfunction, Depression/Anxiety, and Workload Perception
Tiredness on waking, poor subjective sleep quality (C1) and daytime dysfunction (C7) correlated positively with each other (Table 2). PHQ-4 (depression/anxiety) correlated positively with Tiredness on waking, poor subjective sleep quality (C1) and daytime dysfunction (C7) on workdays and work-free days (Figure 1).
Figure 1.
Relationship between depression/anxiety and self-reported sleep variables. Depression/anxiety (PHQ-4) and (a) tiredness on waking, (b) subjective sleep quality (C1), (c) daytime dysfunction (C7).
Note: n=67, **P<0.01, ***P<0.001, ****P<0.0001, green symbols semi-transparent. Dashed lines, 95% confidence intervals.
Abbreviations: C1, component 1 of PSQI; C7, component 7 of PSQI; PHQ-4, Patient Health Questionnaire-4; PSQI Pittsburgh Sleep Quality Index; r, Spearman correlation coefficient.
Perception of workload correlated positively with tiredness on waking and poor subjective sleep quality (C1) on workdays (Figure 2a and b), and with daytime dysfunction (C7) on workdays and work-free days (Figure 2c).
Figure 2.
Relationship between workload perception and self-reported sleep variables. Workload and (a) tiredness on waking, (b) subjective sleep quality (C1), (c) daytime dysfunction (C2).
Note: n=67, *P<0.05; **P<0.01, ***P<0.001, ****P<0.0001, green symbols semi-transparent. Dashed lines, 95% confidence intervals.
Abbreviations: C1, component 1 of PSQI; C7, component 7 of PSQI; PSQI; Pittsburgh Sleep Quality Index; r, Spearman correlation coefficient.
Associations Between Wearable-Measured and Self-Reported Sleep Variables
First, we investigated the relationship between objective measures for sleep duration with self-reported sleep quality. The difference in wearable-measured sleep duration between workdays and work-free days (Δsleep duration) correlated negatively with tiredness on waking (Figure 3a), global PSQI scores (Figure 3b) and poor subjective sleep quality (C1) (Figure 3c).
Figure 3.
Relationship between the difference in wearable-measured sleep duration between workdays and work-free days (Δsleep duration) with self-reported sleep variables and REM sleep. Relationship between Δsleep duration and (a) tiredness on waking, (b) global PSQI, (c) Subjective sleep quality (C), (d) proportion of REM sleep, (e) the difference in REM sleep phase between workdays and work-free days (Δphase REM sleep).
Note: n=67, *P<0.05; **P<0.01, ***P<0.001, green symbols semi-transparent. Dashed lines, 95% confidence intervals.
Abbreviations: C1, component 1 of PSQI; PSQI, Pittsburgh Sleep Quality Index; r, Spearman correlation coefficient; REM, rapid eye movement.
Next, we investigated the relationship between (Δsleep duration) and wearable-measured REM sleep proportion and the difference in REM sleep phase between workdays and work-free days (Δphase REM sleep). Δsleep duration correlated positively with REM sleep proportion (Figure 3d) on work-free days. Moreover, Δsleep duration correlated positively with the difference in the phase of REM sleep between workdays and work-free days (Δphase REM sleep) (Figure 3e). There was no significant correlation of Δsleep duration with the proportion (r=−0.027, P=0.83) or the Δphase (r=0.08, P=0.25) of deep sleep.
Wearable-measured sleep efficiency of the entire data collection period (total sleep efficiency) correlated negatively with global PSQI scores on workdays and work-free days (Figure 4a) and poor subjective sleep quality (C1) on work-free days (Figure 4b) and positively with REM sleep proportion on workdays and work-free days (Figure 4c). The wearable-measured proportion of wake after sleep onset of the entire data collection period (total WASO) showed corresponding inverse relationships (Figure 4d–f).
Figure 4.
Relationship between wearable-measured and self-reported sleep variables. Associations between wearable-measured sleep efficiency of the total data collection period and (a) global PSQI, (b) C1 of PSQI, (c) proportion of REM sleep. Associations between WASO proportion of the total data collection period and (d) global PSQI, (e) C1 of PSQI, (f) proportion of REM sleep.
Note: n=67, *P<0.05; **P<0.01, green symbols semi-transparent. Dashed lines, 95% confidence intervals.
Abbreviations: C1, component 1 of PSQI, PSQI, Pittsburgh Sleep Quality Index; r, Spearman correlation coefficient; REM, rapid eye movement, WASO, wake after sleep onset.
There were no significant correlations between lifestyle variables and objectively measured sleep variables (Supplementary Table 3).
Discussion
This study, which combines self-reported and Fitbit wearable-based data, contributes to the understanding of the differences between sleep quality on workdays and work-free, both of which contribute to mental well-being and performance.
In line with earlier studies with a general population25 and a larger cohort with a similar population,24 we also found significantly lower global PSQI scores and thus generally better sleep quality on work-free days than on workdays. The analysis of the PSQI components indicates in particular higher subjective sleep quality (C1), lower sleep latency (C2), higher sleep duration (C2), as well as a lower daytime dysfunction (C7) on work-free days, similar to previous studies.24,25 In addition, tiredness on waking, a variable we have used in previous studies as an indicator for sleep quality,13,24,27 was lower on workfree days. Tiredness on waking and daytime dysfunction were associated with poor subjective sleep quality, consistent with good sleep quality being important for feeling rested and performance. Similar to our previous study,24 perceived workload, tiredness on waking, poor subjective sleep quality, and daytime dysfunction were associated with symptoms of depression and anxiety. This highlights the role of work-related psychological stress and subjective sleep quality for mental health and performance, which in turn are directly related to general well-being.46 A meta-analysis demonstrates the negative impact of occupational stress on self-reported sleep quality measured by the standard PSQI47 which primarily reflects sleep quality on workdays.25 By differentiating between workdays and non-workdays, we can conclude, in line with our previous observations,24,27 that perceived workload does indeed have a stronger impact on subjective sleep quality on workdays. However, it appears to have a sustained impact on daytime dysfunction on all days. Thus, reducing work-related stress, for example through meditation,47 could help improve sleep on workdays and the feeling of functionality.
The inclusion of objective measures to assess sleep is pivotal in understanding the foundations of restful sleep. Consistent with our earlier study in young adults,27 the Fitbit-measured duration of sleep and sleep stages (including WASO) were longer on work-free days than on workdays. This indicates a sleep loss on workdays and a sleep compensation on work-free days. Moreover, the phase of sleep and sleep stages (including WASO) were later on work-free days, consistent with social jet lag. In addition, the Fitbit-measured sleep latency was shorter on work-free days, suggesting a higher sleep pressure. Fitbit-measured WASO was also lower on work-free days and consequently sleep efficiency was higher, indicating a better sleep consolidation. Interestingly, the proportion of REM sleep, which was inversely associated with sleep efficiency and the proportion of WASO, was higher on work-free days. This suggests that better sleep consolidation on work-free days is in favor of REM sleep, which appears to be important for sleep quality.27,29,30
A longer sleep duration on work-free days than on workdays was associated with less tiredness on waking, lower global PSQI score, thus self-reported sleep quality,30 and in particular with subjective sleep quality (C1) on work-free days. This suggests an association between objective sleep quantity and subjectively more restful sleep and better sleep quality. This is consistent with sleep quantity and quality being interrelated components of restful sleep.9 These associations also indicate that compensating for work-related sleep loss contributes to subjectively more restful and better sleep quality on work-free days. Sleep compensation appears to be particularly48 but not exclusively (this study) important for late chronotypes. Although official sources appreciate the negative effects of sleep loss on health and sleeping in on weekends as a common way to catch up on sleep and/or return to a normal sleep rhythm, it is often recommended to set the alarm even on weekends for consistency (eg).49 However, since our studies suggest that sleeping in on work-free days contributes to better sleep and thus well-being, the start of work should rather be adjusted to individual needs whenever possible.
Moreover, a longer sleep duration on work-free days than on workdays was associated with a higher proportion of REM sleep. This is consistent with REM sleep, which occurs predominantly in the second half of the night and early in the morning,31,32 is shortened by being woken up early on workdays and prolonged by sleeping in on non-workdays. Similarly, a longer sleep duration on work-free days than on workdays was associated with a later phase of REM sleep. The phase of the REM sleep appears to be strongly controlled by the circadian component of sleep timing17 and is therefore also discussed as a good indicator for the intrinsic phase position.50
We found associations between subjective sleep quality and objective measures for sleep quality in particular on work-free days. Higher Fitbit-measured sleep efficiency was associated with less tiredness on waking, lower global PSQI and C1 scores. Similarly, a lower WASO proportion, rather than WASO duration, was associated with less tiredness on waking, lower global PSQI and C1 scores. This is consistent with the associated between perception of insomnia and restless sleep with critical transitions in the sleep architecture.51 Thus, sleep efficiency and WASO proportion, which can be measured easily using a variety of commercial wearable sleep-tracking devices39,48 appear to be good digital markers for sleep quality, in particular in the absence of workday constraints.
Finally, we would like to emphasize that this study was conducted on healthy, younger adults from a metropolitan area who do not perform shift work. Different results are likely to be obtained in a rural population, older adults, people with chronic illnesses, or shift workers.
Limitations
Our cohort is relatively small and does not represent a cross-section of the population. The Fitbit’s sleep stage analysis is not as accurate as polysomnography. Furthermore, correlations provide a measure of strength and direction of relationships but not causality.
Conclusion
Our data indicate that the feeling of sleeping well and waking up rested contributes to psychological well-being and performance. The poorer sleep quality on workdays compared to work-free days appears to be associated, in addition to work-related psychological factors, primarily to shorter sleep duration, lower sleep efficiency, and a lower proportion of REM sleep in favor of a higher proportion of WASO. Sleep compensation on work-free days appears to have a positive effect on subjective sleep quality and the proportion of REM sleep. To achieve better sleep quality on workdays, measures to increase stress resistance could be implemented and, if possible, work should be started later. Alternatively, the sleep dept should be compensated for on work-free days.
Acknowledgments
We thank the participants for their data and support for this study.
Funding Statement
Article Processing Charge was funded by the Open-Access-Fund of the Heinrich-Heine University.
Data Sharing Statement
Data will be made available on reasonable request from the corresponding author.
Author Contributions
All authors have accepted responsibility for the entire content of this paper and consented to its submission. All authors have read and agreed to the published version of the paper. All authors agreed on the journal to which the article has been submitted. Conceptualization: CvG; methodology: CvG; formal analysis: CvG; investigation; DA; writing-original draft: CvG; writing-review & editing: CvG, DA, resources: CvG; data curation: CvG; visualization, supervision, project administration: CvG.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Iranzo A. An overview on sleep medicine. Adv Exp Med Biol. 2022;1384:3–15. doi: 10.1007/978-3-031-06413-5_1 [DOI] [PubMed] [Google Scholar]
- 2.Zager A, Andersen ML, Ruiz FS, Antunes IB, Tufik S. Effects of acute and chronic sleep loss on immune modulation of rats. Am J Physiol Regul Integr Comp Physiol. 2007;293(1):R504–R509. doi: 10.1152/ajpregu.00105.2007 [DOI] [PubMed] [Google Scholar]
- 3.Kecklund G, Axelsson J. Health consequences of shift work and insufficient sleep. BMJ. 2016;355. doi: 10.1136/bmj.i5210 [DOI] [PubMed] [Google Scholar]
- 4.Lamon S, Morabito A, Arentson‐Lantz E, et al. The effect of acute sleep deprivation on skeletal muscle protein synthesis and the hormonal environment. Physiol. Rep. 2021;9(1):e14660. doi: 10.14814/phy2.14660 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Killgore WD. Effects of sleep deprivation on cognition. Prog Brain Res. 2010;185:105–129. [DOI] [PubMed] [Google Scholar]
- 6.Triantafillou S, Saeb S, Lattie EG, Mohr DC, Kording KP. Relationship between sleep quality and mood: ecological momentary assessment study. JMIR Mental Health. 2019;6(3):e12613. doi: 10.2196/12613 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nelson KL, Davis JE, Corbett CF. Sleep quality: an evolutionary concept analysis. Nurs Forum. 2022;57(1):144–151. doi: 10.1111/nuf.12659 [DOI] [PubMed] [Google Scholar]
- 8.Sharma S, Kavuru M. Sleep and metabolism: an overview. Int J Endocrinol. 2010;2010:1–12. doi: 10.1155/2010/270832 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kohyama J. Which is more important for health: sleep quantity or sleep quality? Children. 2021;8(7). doi: 10.3390/children8070542 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Scott AJ, Webb TL, Martyn-St James M, Rowse G, Weich S. Improving sleep quality leads to better mental health: a meta-analysis of randomised controlled trials. Sleep Medicine Reviews. 2021;60:101556. doi: 10.1016/j.smrv.2021.101556 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Litwiller B, Snyder LA, Taylor WD, Steele LM. The relationship between sleep and work: a meta-analysis. J Appl Psychol. 2017;102(4):682–699. doi: 10.1037/apl0000169 [DOI] [PubMed] [Google Scholar]
- 12.Pilcher JJ, Ginter DR, Sadowsky B. Sleep quality versus sleep quantity: relationships between sleep and measures of health, well-being and sleepiness in college students. J Psychosom Res. 1997;42(6):583–596. doi: 10.1016/s0022-3999(97)00004-4 [DOI] [PubMed] [Google Scholar]
- 13.von Gall C, Muth T, Angerer P. Sleep duration on workdays is correlated with subjective workload and subjective impact of high workload on sleep in young healthy adults. Brain Sci. 2023;13(5). doi: 10.3390/brainsci13050818 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health. 2015;1(1):40–43. doi: 10.1016/j.sleh.2014.12.010 [DOI] [PubMed] [Google Scholar]
- 15.Meyer N, Harvey AG, Lockley SW, Dijk DJ. Circadian rhythms and disorders of the timing of sleep. Lancet. 2022;400(10357):1061–1078. doi: 10.1016/S0140-6736(22)00877-7 [DOI] [PubMed] [Google Scholar]
- 16.Borbely AA, Daan S, Wirz-Justice A, Deboer T. The two-process model of sleep regulation: a reappraisal. J Sleep Res. 2016;25(2):131–143. doi: 10.1111/jsr.12371 [DOI] [PubMed] [Google Scholar]
- 17.Dijk DJ, von Schantz M. Timing and consolidation of human sleep, wakefulness, and performance by a symphony of oscillators. J Biol Rhythms. 2005;20(4):279–290. doi: 10.1177/0748730405278292 [DOI] [PubMed] [Google Scholar]
- 18.Korf H-W, von Gall C. Circadian Physiology. In: Pfaff DW, Volkow ND, editors. Neuroscience in the 21st Century. 3. Springer Science+Business Media; 2021. [Google Scholar]
- 19.Lack L, Bailey M, Lovato N, Wright H. Chronotype differences in circadian rhythms of temperature, melatonin, and sleepiness as measured in a modified constant routine protocol. Nat. Sci. Sleep. 2009;1–8. doi: 10.2147/NSS.S6234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Roenneberg T, Wirz-Justice A, Merrow M. Life between clocks: daily temporal patterns of human chronotypes. J. Biol. Rhythms. 2003;18(1):80–90. doi: 10.1177/0748730402239679 [DOI] [PubMed] [Google Scholar]
- 21.Montaruli A, Castelli L, Mule A, et al. Biological rhythm and chronotype: new perspectives in health. Biomolecules. 2021;11(4). doi: 10.3390/biom11040487 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Vitale JA, Roveda E, Montaruli A, et al. Chronotype influences activity circadian rhythm and sleep: differences in sleep quality between weekdays and weekend. Chronobiol Int. 2015;32(3):405–415. doi: 10.3109/07420528.2014.986273 [DOI] [PubMed] [Google Scholar]
- 23.Roenneberg T, Pilz LK, Zerbini G, Winnebeck EC. Chronotype and Social Jetlag: a (Self-) Critical Review. Biology. 2019;8(3). doi: 10.3390/biology8030054 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Aslamyar D, Pilz LK, von Gall C. Relationships between self-reported sleep quality, quantity and timing on workdays vs work-free days and lifestyle factors in healthy adults. Nat Sci Sleep. 2025;17:1641–1654. doi: 10.2147/NSS.S537593 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pilz LK, Keller LK, Lenssen D, Roenneberg T. Time to rethink sleep quality: PSQI scores reflect sleep quality on workdays. Sleep. 2018;41(5). doi: 10.1093/sleep/zsy029 [DOI] [PubMed] [Google Scholar]
- 26.Pilz LK, Levandovski R, Oliveira MAB, Hidalgo MP, Roenneberg T. Sleep and light exposure across different levels of urbanisation in Brazilian communities. Sci Rep. 2018;8(1):11389. doi: 10.1038/s41598-018-29494-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.von Gall C, Holub L, Pfeffer M, Eickhoff S. Chronotype-dependent sleep loss is associated with a lower amplitude in circadian rhythm and a higher fragmentation of rem sleep in young healthy adults. Brain Sci. 2023;13(10). doi: 10.3390/brainsci13101482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Physiology, Sleep Stages. StatPearls Publishing; 2023. [PubMed] [Google Scholar]
- 29.Barbato G. REM sleep: an unknown indicator of sleep quality. Int J Environ Res Public Health. 2021;18(24). doi: 10.3390/ijerph182412976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193–213. doi: 10.1016/0165-1781(89)90047-4 [DOI] [PubMed] [Google Scholar]
- 31.Williams RL, Agnew HW, Webb WB. Sleep patterns in young adults: an EEG study. Electroencephalogr Clin Neurophysiol. 1964;17:376–381. doi: 10.1016/0013-4694(64)90160-9 [DOI] [PubMed] [Google Scholar]
- 32.Williams RL, Agnew HW Jr, Webb WB. Sleep patterns in the young adult female: an EEG study. Electroencephalogr Clin Neurophysiol. 1966;20(3):264–266. doi: 10.1016/0013-4694(66)90092-7 [DOI] [PubMed] [Google Scholar]
- 33.von Gall C, Holub L, Ali AAH, Eickhoff S. Timing of Deep and REM Sleep Based on Fitbit Sleep Staging in Young Healthy Adults under Real-Life Conditions. Brain Sci. 2024;14(3). doi: 10.3390/brainsci14030260 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wulterkens BM, Fonseca P, Hermans LWA, et al. It is all in the wrist: wearable sleep staging in a clinical population versus reference polysomnography. Nat Sci Sleep. 2021;13:885–897. doi: 10.2147/NSS.S306808 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Cook JD, Eftekari SC, Dallmann E, Sippy M, Plante DT. Ability of the Fitbit Alta HR to quantify and classify sleep in patients with suspected central disorders of hypersomnolence: a comparison against polysomnography. J Sleep Res. 2018;28:e12789. doi: 10.1111/jsr.12789 [DOI] [PubMed] [Google Scholar]
- 36.de Zambotti M, Goldstone A, Claudatos S, Colrain IM, Baker FC. A validation study of Fitbit Charge 2 compared with polysomnography in adults. Chronobiol Int. 2018;35(4):465–476. doi: 10.1080/07420528.2017.1413578 [DOI] [PubMed] [Google Scholar]
- 37.Haghayegh S, Khoshnevis S, Smolensky MH, Diller KR, Castriotta RJ. Accuracy of wristband fitbit models in assessing sleep: systematic review and meta-analysis. J Med Internet Res. 2019;21(11):ARTNe16273. doi: 10.2196/16273 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Liang Z, Chapa-Martell MA. A multi-level classification approach for sleep stage prediction with processed data derived from consumer wearable activity trackers. Front Digit Health. 2021;3:665946. doi: 10.3389/fdgth.2021.665946 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Chinoy ED, Cuellar JA, Jameson JT, Markwald RR. Performance of four commercial wearable sleep-tracking devices tested under unrestricted conditions at home in healthy young adults. Nat Sci Sleep. 2022;14:493–516. doi: 10.2147/NSS.S348795 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lee YJ, Lee JY, Cho JH, Kang YJ, Choi JH. Performance of consumer wrist-worn sleep tracking devices compared to polysomnography: a meta-analysis. J Clin Sleep Med. 2025;21(3):573–582. doi: 10.5664/jcsm.11460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zheng NS, Annis J, Master H, et al. Sleep patterns and risk of chronic disease as measured by long-term monitoring with commercial wearable devices in the all of us research program. Nat Med. 2024;30(9):2648–2656. doi: 10.1038/s41591-024-03155-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Löwe B, Wahl I, Rose M, et al. A 4-item measure of depression and anxiety: validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population. J Affective Disorders. 2010;122(1–2):86–95. doi: 10.1016/j.jad.2009.06.019 [DOI] [PubMed] [Google Scholar]
- 43.Fitbit. Available from: https://support.google.com/fitbit/answer/14237111?sjid=12136253115896863514-EU#exp. Accessed December 09, 2025.
- 44.Fitbit. Available from: https://help.fitbit.com/articles/en_US/Help_article/2439.htm. Accessed December 09, 2025.
- 45.WHO. Available from: https://www.who.int/europe/news-room/fact-sheets/item/a-healthy-lifestyle-who-recommendations. Accessed December 09, 2025.
- 46.Zhu Y, Meng R, Jiang C, et al. Sleep quality and subjective well-being in healthcare students: examining the role of anxiety and depression. Front Public Health. 2023;11:1281571. doi: 10.3389/fpubh.2023.1281571 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Mao Y, Raju G, Zabidi MA. Association between occupational stress and sleep quality: a systematic review. Nat Sci Sleep. 2023;15:931–947. doi: 10.2147/NSS.S431442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Vitale JA, Calogiuri G, Weydahl A. Influence of chronotype on responses to a standardized, self-paced walking task in the morning vs afternoon: a pilot study. Percept Mot Skills. 2013;116(3):1020–1028. doi: 10.2466/06.19.PMS.116.3.1020-1028 [DOI] [PubMed] [Google Scholar]
- 49.Newsom R. Sleep Debt: The Hidden Cost of Insufficient Rest. Available from: https://www.sleepfoundation.org/how-sleep-works/sleep-debt-and-catch-up-sleep. 2025. July 09Accessed December 09, 2025.
- 50.Horvath CG, Schneider B, Rozner B, Koczur M, Bodizs R. Interrelationships between sleep quality, circadian phase and rapid eye movement sleep: deriving chronotype from sleep architecture. Behav Res Methods. 2025;57(5):150. doi: 10.3758/s13428-025-02671-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Jacobs J, Martin CE, Fuemmeler B, Chen S. Profiling the sleep architecture of ageing adults using a seven-state continuous-time Markov model. J Sleep Res. 2025;34(2):e14331. doi: 10.1111/jsr.14331 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data will be made available on reasonable request from the corresponding author.